Measuring symptoms and wellness in the multiple sclerosis patient: issues in measurement

Appropriate measurement of any state of health or disease depends on several factors. First and foremost is the scientific question to be answered, which can range from basic science to clinical processes to societal impact. Questions primarily address either the current state of the patient or are designed to measure change over time. Although it is sometimes possible to accomplish both objectives with one set of questions, this can be difficult and may not be appropriate. Questions must be designed appropriately to ensure reliability of the answer, validity of the concept being measured and the sensitivity or responsiveness of the measure to change. It is also becoming standard to evaluate all parameters for the clinical (or scientific) relevance of the measured value or change in value over time. The design of a measurement tool must consider the type of data to be collected, the difficulties in collecting complete data and the plan for data analysis. Ever-increasing computing power has provided novel approaches to analysis allowing for a broader range of measurement instruments. This review examines some important basic principles of measurement and considers issues as they apply to examples of measures used for patients with multisymptomatic illnesses such as multiple sclerosis, with a particular focus on application of the spasticity 0-10 Numerical Rating Scale and its correlation with the Patient Global Impression of Change scale.